论文部分内容阅读
The aim of our study was to estimate the relationship between growth (leaf length, leaf width, leaf weight, plant height, plant width, plant weight) and functional components (glucosinolate, anthocyanin, amino acid, sugar, and carotenoid), and spectral reflectance of Chinese cabbage leaf using five prediction methods.Chinese cabbage leaf samples used in this study were grown in a plant factory.Spectral reflectance data of leaf at wavelength bands from 190 nm to 1,130 nm were collected by using a UV/VIS/NIR spectrometer in a dark condition.Scans were repeated 12 times evenly in the blade part for each leaf measurement.For determination of important wavelengths, correlation coefficient spectrum, mathematical transformations (eR, R2,l/R, Ln (R),错误!未找到引用源.),partial least square regression (PLSR), and stepwise multiple linear regression (SMLR) were applied.Performances of SMLR and PLSR procedures were good.Most of good predictive model was obtained by using SMLR procedure and the square root of reflectance data for functional components and physical parameters.The wavelengths at approximately 580 and 650 nm were identified by both PLSR and SMLR analyses.These two wavelengths have been reported that related to estimating anthocyanin, chlorophyll and nitrogen concentrations.These results would be useful to provide design guidance for a multiple-wavelength property sensor usable for precision agriculture and other high-resolution sensing applications.